MEV Meets LPing: Extending an Olive Branch
MEV Meets LPing: Extending an Olive Branch
In the last blog post, we observed how Maximum Extractable Value (MEV) searchers are constantly arbitraging stale prices provided by passive LPs, profiting off the difference between the quoted price on the AMM and the true market price. In traditional market making, this is known as adverse selection — the presence of new information (i.e. hacks, sentiment change, Twitter news, etc.) causes smarter market makers to update their prices and adversely select against the slower market makers. Loss-versus-rebalancing (LVR), a term invented by Roughgarden et al. (2022), outlines a closed form solution that defines the running cost of this adverse selection.
According to MEV-Explore by Flashbots, Uniswap V2 contributes to 63% of all MEV on Ethereum, and Uniswap V3 contributes to 16%. Given that $676M has been extracted thus far, that equates to $540M of MEV on Uniswap across 3 years, or $180M per year. Another source, EigenPhi, has shown that $138M has been extracted from Uniswap V3 from Jan. 1 — Oct. 31, 2022. Correspondingly, the amount of fees generated by Uniswap V3 during this time is ~$600M, so redistributing MEV could improve profitability by ~20%.
In order to repay MEV back to the LPs, there are a few core research directions.
The first mechanism is through an auction of user order flow to MEV searchers, which is being developed by protocols like Rook and DFlow. The profit earned by this auction comes back to the protocol, and can be distributed in a fair way between the LP and trader. The DAO can also choose to offer liquidity mining incentives in terms of its native governance token, which can be used to vote upon giving LP’s a share of the DAO’s treasury that represents the MEV profits.
The second approach is to run the AMM as an app-chain, thereby controlling the MEV at the validation level. For instance, by embedding an oracle into the beginning of each block, or allowing MEV to be captured in a controlled fashion, the underlying L1 can choose where the LVR goes. Current companies in this space include Sei and Duality.
Another approach of redistributing MEV is through dynamic fees that charge informed traders a higher fee on swaps. Dynamic fees also solve the LVR problem by equating the fee to an options premium for an analogous instrument. For instance, since the LP is buying more of the less valuable asset, its payoff reflects that of a short straddle (selling both call and put options), but without collecting a consistent premium. Instead, the premium is reflected as trading fees distributed to LPs. The true value of this premium forgone to the market (in a continuous-time, black-scholes setting) can be clawed back to the LP in the form of dynamic fees during more volatile times. This can be done at the AMM level by designing dynamic fees into the AMM or even by LP management protocols, which may choose to offer its users the ability to exit their position automatically when the premium paid by fees is lower than the price of an option using the implied volatility of the market at any given time via active liquidity management using a collection of keepers.
One of the primary downsides of a few of the approaches mentioned is they either require building a new DEX from the ground up or making a significant change to an existing DEX. The distribution problem of bootstrapping a flywheel that attracts both liquidity providers and traders is a difficult challenge, especially when faced against the robust system that has been developed by Uniswap.
Increasing the composability of the LP position is another way to increase LP profitability. Since an LP position represents a short straddle, a protocol can be created to lend this LP position out. The borrower is then able to short this LP to achieve the highly appealing long straddle (also known as “buying wings”) and pay a funding rate back to the LP. Since the LP position already generates fees, the borrower will likely pay a lower rate than what they may see in the typical options market. Regardless, the market chooses a fair premium for the payoff, reducing the LVR of the LP.
While MEV and LVR present clear concerns to the profitability of LPs, the next blog post outlines a number of ways to still remain profitable through a hedged and sustainable portfolio.
About OpenBlock Labs
OpenBlock is a platform that empowers Web3 organizations with the data intelligence needed to take action and drive growth. From managing grants to optimizing incentive spend, leading DAOs and blockchain networks utilize this technology to maintain billions of dollars in value within their ecosystems.
If you are interested in working with OpenBlock Labs, please visit us at openblocklabs.com or reach out via Twitter at @openblocklabs.
In the last blog post, we observed how Maximum Extractable Value (MEV) searchers are constantly arbitraging stale prices provided by passive LPs, profiting off the difference between the quoted price on the AMM and the true market price. In traditional market making, this is known as adverse selection — the presence of new information (i.e. hacks, sentiment change, Twitter news, etc.) causes smarter market makers to update their prices and adversely select against the slower market makers. Loss-versus-rebalancing (LVR), a term invented by Roughgarden et al. (2022), outlines a closed form solution that defines the running cost of this adverse selection.
According to MEV-Explore by Flashbots, Uniswap V2 contributes to 63% of all MEV on Ethereum, and Uniswap V3 contributes to 16%. Given that $676M has been extracted thus far, that equates to $540M of MEV on Uniswap across 3 years, or $180M per year. Another source, EigenPhi, has shown that $138M has been extracted from Uniswap V3 from Jan. 1 — Oct. 31, 2022. Correspondingly, the amount of fees generated by Uniswap V3 during this time is ~$600M, so redistributing MEV could improve profitability by ~20%.
In order to repay MEV back to the LPs, there are a few core research directions.
The first mechanism is through an auction of user order flow to MEV searchers, which is being developed by protocols like Rook and DFlow. The profit earned by this auction comes back to the protocol, and can be distributed in a fair way between the LP and trader. The DAO can also choose to offer liquidity mining incentives in terms of its native governance token, which can be used to vote upon giving LP’s a share of the DAO’s treasury that represents the MEV profits.
The second approach is to run the AMM as an app-chain, thereby controlling the MEV at the validation level. For instance, by embedding an oracle into the beginning of each block, or allowing MEV to be captured in a controlled fashion, the underlying L1 can choose where the LVR goes. Current companies in this space include Sei and Duality.
Another approach of redistributing MEV is through dynamic fees that charge informed traders a higher fee on swaps. Dynamic fees also solve the LVR problem by equating the fee to an options premium for an analogous instrument. For instance, since the LP is buying more of the less valuable asset, its payoff reflects that of a short straddle (selling both call and put options), but without collecting a consistent premium. Instead, the premium is reflected as trading fees distributed to LPs. The true value of this premium forgone to the market (in a continuous-time, black-scholes setting) can be clawed back to the LP in the form of dynamic fees during more volatile times. This can be done at the AMM level by designing dynamic fees into the AMM or even by LP management protocols, which may choose to offer its users the ability to exit their position automatically when the premium paid by fees is lower than the price of an option using the implied volatility of the market at any given time via active liquidity management using a collection of keepers.
One of the primary downsides of a few of the approaches mentioned is they either require building a new DEX from the ground up or making a significant change to an existing DEX. The distribution problem of bootstrapping a flywheel that attracts both liquidity providers and traders is a difficult challenge, especially when faced against the robust system that has been developed by Uniswap.
Increasing the composability of the LP position is another way to increase LP profitability. Since an LP position represents a short straddle, a protocol can be created to lend this LP position out. The borrower is then able to short this LP to achieve the highly appealing long straddle (also known as “buying wings”) and pay a funding rate back to the LP. Since the LP position already generates fees, the borrower will likely pay a lower rate than what they may see in the typical options market. Regardless, the market chooses a fair premium for the payoff, reducing the LVR of the LP.
While MEV and LVR present clear concerns to the profitability of LPs, the next blog post outlines a number of ways to still remain profitable through a hedged and sustainable portfolio.
About OpenBlock Labs
OpenBlock is a platform that empowers Web3 organizations with the data intelligence needed to take action and drive growth. From managing grants to optimizing incentive spend, leading DAOs and blockchain networks utilize this technology to maintain billions of dollars in value within their ecosystems.
If you are interested in working with OpenBlock Labs, please visit us at openblocklabs.com or reach out via Twitter at @openblocklabs.
